The overall aim of this project is to understand the biochemistry, biology, and pathology of cell surface-associated serine proteases, with emphasis on determining their contribution to the development, regeneration, and malignant transformation of oral tissues. Novel cell surface serine proteases in epidermal development, repair, and malignancy: The type II transmembrane serine proteases (TTSPs) constitute a recently recognized family of membrane-bound serine proteases with largely unknown functions. We have continued our efforts to identify novel TTSPs and determine their contribution to development and malignancy. The TTSP matriptase is expressed with remarkable consistency in human carcinoma, including oral squamous cell carcinoma, and high matriptase expression correlates with poor disease outcome. In collaboration with Silvio Gutkind, MCU, OPCB, we investigated the causal relationship between matriptase overexpression and squamous cell carcinogenesis. Unexpectedly, we found that matriptase possesses a strong oncogenic potential when unopposed by its endogenous inhibitor, HAI-1. Modest overexpression of matriptase in the epidermis of transgenic mice caused spontaneous squamous cell carcinoma and dramatically potentiated carcinogen-induced tumor formation. Matriptase-induced malignancy was preceded by all the stereotypic changes observed in human oral squamous cell carcinogenesis, including progressive hyperplasia, dysplasia, fibrosis, and dermal inflammation. At the molecular level, matriptase expression induced robust activation of the pro-tumorgenic PI3K-Akt signaling pathway, functionally emulated the effects of chemical tumor promoters, and supported both ras-dependent and ras-independent transformation. Increasing epidermal HAI-1 expression to specifically inhibit the proteolytic activity of matriptase completely negated the oncogenic effects of the membrane protease. The data implicate dysregulated matriptase expression in epithelial carcinogenesis and demonstrate that aberrant cell surface proteolysis can suffice to drive malignant transformation. Using bioinformatics, in collaboration with Toni Antalis, University of Maryland, we have identified a third member of the matriptase subfamily of TTSPs, matriptase-3, and have molecularly cloned and biochemically characterized the new cell surface serine protease. The matriptase-3 gene was conserved in all ten vertebrate species examined, and it located to syntenic regions of human mouse, rat, and chicken chromosomes. Bioinformatic analysis - confirmed by direct cDNA cloning - revealed a functional TTSP with 31% amino acid identity with both matriptase and matriptase-2 and an identical domain structure. Matriptase-3 displayed species-conserved expression, with the highest mRNA levels found in brain, skin, reproductive, and oropharyngeal tissues. The matriptase-3 cDNA directed the expression of a 90 kDa N-glycosylated protein that localized to the cell surface, as assessed by cell-surface biotin labeling. The purified activated matriptase-3 serine protease domain hydrolyzed synthetic peptide substrates with a preference for Arg at position P(1), and showed proteolytic activity towards several macromolecular substrates. Interestingly, activated matriptase-3 formed stable inhibitor complexes with an array of serpins, lending support to our previous proposition that TTSPs may be novel targets for serpin inhibition. The study identifies matriptase-3 as an evolutionarily conserved, biologically active TTSP of the matriptase subfamily. Urokinase receptor-associated protein and intracellular collagen turnover: The urokinase plasminogen activator (uPA) receptor (uPAR)-associated protein (uPARAP) is a member of the macrophage mannose receptor family of transmembrane glycoproteins that was first identified as a protein forming a tri-molecular complex with uPA and uPAR. We recently reported the unexpected finding that uPARAP is critical for the cellular uptake and lysosomal degradation of a range of collagen species by cultured fibroblasts. Studies conducted with Alfredo Molinolo, OPCB and Dr. Niels Behrendt, University of Copenhagen revealed that uPARAP expression is selectively induced in mesenchymal cells embedded within the reactive collagen-rich stroma of several human carcinomas, including oral squamous cell carcinoma and salivary gland tumors. This suggested a possible role of the receptor in collagen turnover during tumor invasion. To test this hypothesis, in collaboration with Dr. Niels Behrendt, University of Copenhagen, we generated carcinoma-prone uPARAP-deficient and -sufficient mice using the validated MMTV-Polyomavirus middle T model of mammary adenosquamous carcinoma. Interestingly, the genetic ablation of uPARAP impaired collagen turnover critical to tumor expansion, as evidenced by the abrogation of cellular collagen uptake, tumor fibrosis, and blunted tumor growth. These studies identified uPARAP as a key mediator of collagen turnover in a pathophysiological context and showed that intracellular collagen degradation is a functionally relevant pathway for matrix turnover during tumor progression. As MMP and uPARAP-dependent pathways of collagen degradation may be at least partially independent, the data raise the intriguing possibility that pharmacological inhibition of MMP activity aimed at preventing pathological connective tissue destruction during cancer and other tissue destructive disease, may be functionally counteracted by increased uPARAP-dependent intracellular collagen degradation. Reengineered bacterial cytotoxins as antitumor and protease imaging agents: We have continued our long-standing collaboration with Stephen Leppla, MPS, NIAID, on the development of cell surface protease-activated bacterial cytotoxins as therapeutic agents for cancer. Our previous work provided ?proof of concept, by showing that cell surface uPA-activated reengineered anthrax toxin displays limited toxicity to normal tissues, but broad and potent uPA-dependent tumorcidal activity in mice. We have worked via a CRADA with the biopharmaceutical company OncoTac and Arthur Frankel, Scott & White, Temple, Texas on the clinical development of uPA activated anthrax toxins. In complementary studies, we have worked on new ways to further increase the tumor therapeutic index of modified bacterial cytotoxins. In this respect, we devised a method for making the activation of anthrax toxin dependent on two distinct tumor-associated proteases. Anthrax toxin protective antigen (PrAg) requires site-specific endoproteolytic cleavage to assemble into a functional heptamer on which the binding site for lethal factor (LF), the catalytic moiety of anthrax toxin, spans two adjacent monomers. These features of PrAg suggested that high cell type specificity in tumor targeting could be obtained using mutated PrAg monomers that would generate functional LF binding sites only through intermolecular complementation and were activated by two different tumor-associated proteases. PrAg mutants were constructed that contained mutations in different LF-binding subsites of the heptameric PrAg. These PrAg mutants were further reengineered to be cleaved by either uPA or MMPs. The individual PrAg mutants had very low toxicity due to greatly decreased LF binding. However, when administered to tumor cells that contained both uPA and MMP, they assembled into functional heptamers with LF binding sites that displayed dissociation constants for LF approaching wildtype PrAg. The two complementing PrAg proteins had greatly reduced toxicity to mice, and an increased therapeutic index relative to uPA activated toxins, which made them very effective in the treatment of transplanted tumors when administered either locally or systemically.